Error -10846 during multichannel, single scan AI

[Wohlsbacher]

Hello,

I want to analyse the frequency response of a circuit using a 6052E-DAQ board and LabVIEW (System: WIN98, Pentium 2, 128MB RAM).
For this purpose I first start a continuous analog output and then try to read four AI channels simultaneously (different input limits).
I thought AI Config should allocate enough buffer to hold one full reading (AI Start should fill one buffer by default), but an overwrite error indicates that the buffer is used circularly. After all I do not get why?

I would be pretty grateful if someone could tell me what LabVIEW does and why I get this error. After all it doesn't occur every time but I'd prefer to never get it.

My settings are:
Samples/CH = 8196
Scan rate = 50 kHz


Thanks for your help

Dirk Neumann

[Wohlsbacher]

After all the expression "single scan AI" in my header might be confusing (well it's just wrong). What I meant was a single buffered AI.

But I still cann't get why LabVIEW produces an overwrite error when I want to fill only one buffer.

Greetings

Dirk Neumann

[Andrew Madry]

Dirk
I've noticed that irrespective of the size of buffer one sets that the famous -10846 error can occur and as you say it's not consistent. It seems to happen at the start of a continuous acquisition (which is what you seem to be experiencing). I've written programs to log when it happens. The strange thing is that when you monitor the scan backlog of AI read there hasn't been an overflow. I use an DAQ Card AI-16XE-50.
What I sometimes do is ignore the -10846 error if the data is just going to a graph display where it doesn't matter too much if a block is dropped. The attached vi makes the error a warning so it won't abort any following vi's and the program will keep running. However in the case that it is essential that no data is dropped, for e
xample if you are saving data to a file, you want to put this vi in a case statement so you can trap the error and stop the program.
regards
Andrew Madry
ps dBVac.vi wasn't there but the vi looked OK.

[Wohlsbacher]

Hey Andrew,

first of all thanks for your answer. Unfortunately I cann't open your VI (LabVIEW 6.0.2), but I can imagine the idea. BTW, I don't want to do a continuous acquisition but a 4-channel buffered (4096 points) analog input.
However I need my data. I use a for-loop to step through the frequencies (changing the analog output update rate) and use the aquired input data to calculate the response at each frequency.

But well, it should be fixable.

Joyous greeting

Dirk